[Issue]: torch-rocm for windows on 7600xt seems broken #4528
Description
Issue Description
After an update a while back when I try to generate an image what is shown in the log occurs.
I do not know any more than that, the help I've found on discord seems to think it may be a compatibility issue with Rocm but if so I don't know how to test that. I have also tried things like deleting the Venv file and a complete clean install of Sdnext. Nothing seems to work.
Version Platform Description
Firefox
Python: version=3.12.0
Device Name Normalfella112
Processor Intel(R) Core(TM) i7-8700 CPU @ 3.20GHz 3.19 GHz
Installed RAM 64.0 GB
Storage 224 GB SSD GIGABYTE GP-GSTFS31240GNTD, 932 GB HDD WDC WD10EZEX-08WN4A0
Graphics Card AMD Radeon RX 7600 XT (16 GB)
Device ID
Product ID
System Type 64-bit operating system, x64-based processor
Pen and touch No pen or touch input is available for this display
Relevant log output
Using VENV: E:\Anewforge\SDNEXT\sdnext\venv
09:59:31-254590 INFO Starting SD.Next
09:59:31-260575 INFO Logger: file="E:\Anewforge\SDNEXT\sdnext\sdnext.log" level=DEBUG host="Normalfella112" size=82
mode=create
09:59:31-263566 INFO Python: version=3.12.0 platform=Windows
bin="E:\Anewforge\SDNEXT\sdnext\venv\Scripts\python.exe"
venv="E:\Anewforge\SDNEXT\sdnext\venv"
09:59:31-480572 INFO Version: app=sd.next updated=2026-01-01 commit=56a8aeaea branch=master
url=https://github.com/vladmandic/sdnext/tree/master kanvas=main ui=main
09:59:32-140075 TRACE Repository branches: active=master available=['dev', 'master', 'upstream']
09:59:32-425006 INFO Version: app=sd.next latest=2026-01-01T09:35:43Z hash=56a8aeae branch=master
09:59:32-453825 INFO Platform: arch=AMD64 cpu=Intel64 Family 6 Model 158 Stepping 10, GenuineIntel system=Windows
release=Windows-10-10.0.19045-SP0 python=3.12.0 locale=('English_United States', '1252')
docker=False
09:59:32-459407 DEBUG Packages: prefix=venv site=['venv', 'venv\\Lib\\site-packages']
09:59:32-461402 INFO Args: ['--use-rocm']
09:59:32-462400 DEBUG Setting environment tuning
09:59:32-464394 DEBUG Torch allocator: "garbage_collection_threshold:0.80,max_split_size_mb:512"
09:59:32-466389 INFO Verifying torch installation
09:59:32-471552 DEBUG Torch overrides: cuda=False rocm=True ipex=False directml=False openvino=False zluda=False
09:59:32-741960 INFO ROCm: agents=['gfx1102']
09:59:32-743695 INFO ROCm: version=7.2, using agent gfx1102
09:59:32-745691 INFO Python: version=3.12.0 platform=Windows
bin="E:\Anewforge\SDNEXT\sdnext\venv\Scripts\python.exe"
venv="E:\Anewforge\SDNEXT\sdnext\venv"
09:59:32-754698 INFO ROCm: HSA_OVERRIDE_GFX_VERSION config overridden: device=gfx1102 version=11.0.0
09:59:58-786192 INFO Torch backend: version="2.9.1+rocm7.10.0a20251120" type=ROCm HIP=7.2.0-c0adf921a0
09:59:58-790177 INFO Torch detected: gpu="AMD Radeon RX 7600 XT" vram=16368 arch=(11, 0) cores=16
09:59:58-966363 INFO Install: verifying requirements
09:59:59-015585 DEBUG Timestamp repository update time: Thu Jan 1 03:35:43 2026
09:59:59-017580 INFO Startup: standard
09:59:59-019798 INFO Verifying submodules
10:00:03-247814 DEBUG Git submodule: extensions-builtin/sd-extension-chainner / main
10:00:03-330326 DEBUG Git submodule: extensions-builtin/sd-extension-system-info / main
10:00:03-407322 DEBUG Git submodule: extensions-builtin/sdnext-kanvas / main
10:00:03-520389 DEBUG Git detached head detected: folder="extensions-builtin/sdnext-modernui" reattach=main
10:00:03-522370 DEBUG Git submodule: extensions-builtin/sdnext-modernui / main
10:00:03-596113 DEBUG Git submodule: extensions-builtin/stable-diffusion-webui-rembg / master
10:00:03-707102 DEBUG Git detached head detected: folder="wiki" reattach=master
10:00:03-710093 DEBUG Git submodule: wiki / master
10:00:03-847801 DEBUG Installed packages: 252
10:00:03-849767 DEBUG Extensions all: ['sd-extension-chainner', 'sd-extension-system-info', 'sdnext-kanvas',
'sdnext-modernui', 'stable-diffusion-webui-rembg']
10:00:04-212797 DEBUG Extension installer: builtin=True
file="E:\Anewforge\SDNEXT\sdnext\extensions-builtin\stable-diffusion-webui-rembg\install.py"
10:00:04-301853 DEBUG Extensions all: []
10:00:04-303818 INFO Extensions enabled: ['sd-extension-chainner', 'sd-extension-system-info', 'sdnext-kanvas',
'sdnext-modernui', 'stable-diffusion-webui-rembg']
10:00:04-400033 INFO Install: verifying requirements
10:00:04-406048 DEBUG Setup complete without errors: 1767542404
10:00:04-410187 DEBUG Extension preload: {'extensions-builtin': 0.0, 'extensions': 0.0}
10:00:04-415144 INFO Installer time: total=39.13 torch=26.32 git=4.87 submodules=4.74 latest=1.16 branch=0.54
base=0.32 version=0.21 requirements=0.19 installed=0.19
10:00:04-421896 INFO Command line args: ['--use-rocm'] use_rocm=True args=[]
10:00:04-424084 DEBUG Env flags: []
10:00:04-425108 DEBUG Starting module: <module 'webui' from 'E:\\Anewforge\\SDNEXT\\sdnext\\webui.py'>
10:00:04-455004 WARNING Loader: torch is not built with distributed support
10:00:15-895272 INFO Torch: torch==2.9.1+rocm7.10.0a20251120 torchvision==0.24.0+rocm7.10.0a20251120
10:00:15-900304 INFO Packages: diffusers==0.37.0.dev0 transformers==4.57.3 accelerate==1.12.0 gradio==3.43.2
pydantic==1.10.21 numpy==2.1.2 cv2==4.12.0
10:00:16-785777 DEBUG ONNX: version=1.23.2, available=['AzureExecutionProvider', 'CPUExecutionProvider']
10:00:16-945136 DEBUG State initialized: id=2305959745808
10:00:17-156054 INFO Device detect: memory=16.0 default=balanced
10:00:17-158719 DEBUG Triton: pass=False fn=<module>:has_triton time=0.00
10:00:17-164755 DEBUG Read: file="E:\Anewforge\SDNEXT\sdnext\config.json" json=20 bytes=1178 time=0.000
fn=<module>:load
10:00:17-170616 WARNING Setting validation: unknown=['chainner_models_path']
10:00:17-172955 INFO Engine: backend=Backend.DIFFUSERS compute=rocm device=cuda attention="Scaled-Dot-Product"
mode=no_grad
10:00:17-176944 DEBUG Read: file="html\reference.json" json=152 bytes=72979 time=0.000
fn=_call_with_frames_removed:<module>
10:00:17-180045 DEBUG Torch attention: type="sdpa" kernels=['Flash', 'Memory', 'Math'] overrides=[]
10:00:17-183008 DEBUG Torch attention installed: flashattn=False sageattention=False
10:00:17-188954 DEBUG Torch attention status: flash=False flash3=False aiter=False sage=False flex=True npu=False
xla=False xformers=False
10:00:17-276864 DEBUG Triton: pass=False fn=<module>:set_cuda_params time=0.00
10:00:17-281209 INFO Torch parameters: backend=rocm device=cuda config=Auto dtype=torch.bfloat16 context=no_grad
nohalf=False nohalfvae=False upcast=False deterministic=False tunable=[False, True] fp16=fail
bf16=pass triton=fail optimization="Scaled-Dot-Product"
10:00:17-296173 DEBUG Quantization: registered=SDNQ
10:00:17-298167 INFO Device: device=AMD Radeon RX 7600 XT n=1 hip=7.2.0-c0adf921a0
10:00:17-486299 DEBUG Entering start sequence
10:00:17-489291 DEBUG Initializing
10:00:17-491286 DEBUG Read: file="metadata.json" json=7 bytes=8274 time=0.001 fn=initialize:init_metadata
10:00:17-495276 DEBUG Read: file="cache.json" json=2 bytes=220 time=0.001 fn=initialize:init_cache
10:00:17-548864 INFO Available VAEs: path="models\VAE" items=0
10:00:17-550858 INFO Available UNets: path="models\UNET" items=0
10:00:17-552853 INFO Available TEs: path="models\Text-encoder" items=0
10:00:17-556259 INFO Available Models: safetensors="models\Stable-diffusion":2 diffusers="models\Diffusers":0
reference=152 items=2 time=0.00
10:00:17-574119 INFO Available LoRAs: path="models\Lora" items=3 folders=2 time=0.00
10:00:17-595579 INFO Available Styles: path="models\styles" items=288 time=0.02
10:00:17-689161 INFO Available Detailer: path="models\yolo" items=14 downloaded=0
10:00:17-692644 DEBUG Extensions: disabled=['sdnext-modernui']
10:00:17-694639 INFO Load extensions
10:00:18-022585 DEBUG Extensions init time: total=0.33 stable-diffusion-webui-rembg=0.14
10:00:18-414134 DEBUG Read: file="html/upscalers.json" json=4 bytes=2672 time=0.000 fn=__init__:__init__
10:00:18-417049 DEBUG Read: file="extensions-builtin\sd-extension-chainner\models.json" json=25 bytes=2830
time=0.000 fn=__init__:find_scalers
10:00:18-420860 DEBUG Available chaiNNer: path="E:\Anewforge\SDNEXT\Sdnext folders\modelschaiNNer" defined=25
discovered=0 downloaded=0
10:00:18-429865 INFO Available Upscalers: items=76 downloaded=0 user=0 time=0.40 types=['None', 'Resize', 'Latent',
'AsymmetricVAE', 'WanUpscale', 'DCC', 'VIPS', 'ChaiNNer', 'AuraSR', 'ESRGAN', 'RealESRGAN',
'SCUNet', 'Diffusion', 'SeedVR', 'SwinIR']
10:00:18-491350 INFO Networks: type="video" engines=13 models=67 errors=0 time=0.05
10:00:18-497486 INFO Huggingface: transfer=rust parallel=True direct=False token="None" cache="models\huggingface"
init
10:00:18-509389 WARNING Cache location changed: previous="C:\Users\norma\.cache\huggingface\hub" size=3839 MB
10:00:18-513084 DEBUG Huggingface: cache="models\huggingface" size=0 MB
10:00:18-515079 DEBUG UI start sequence
10:00:18-517074 DEBUG UI image support: kanvas=main
10:00:18-528268 INFO UI locale: name="Auto"
10:00:18-530262 INFO UI theme: type=Standard name="black-teal" available=14
10:00:18-533254 DEBUG UI theme: css="E:\Anewforge\SDNEXT\sdnext\javascript\black-teal.css" base="['sdnext.css',
'timesheet.css']" user="None"
10:00:18-541217 DEBUG UI initialize: tab=txt2img
10:00:18-571140 DEBUG Read: file="html\reference.json" json=152 bytes=72979 time=0.000 fn=list_items:list_reference
10:00:18-601996 DEBUG Networks: type="reference" items={'total': 152, 'ready': 0, 'hidden': 0, 'experimental': 0,
'base': 96, 'distilled': 18, 'quantized': 19, 'community': 15, 'cloud': 2}
10:00:18-614585 DEBUG Networks: type="model" items=152 subfolders=8 tab=txt2img folders=['models\\Stable-diffusion',
'models\\Reference', 'E:\\Anewforge\\SDNEXT\\sdnext\\models\\Stable-diffusion'] list=0.05
thumb=0.00 desc=0.00 info=0.00 workers=12
10:00:18-622132 DEBUG Networks: type="lora" items=3 subfolders=1 tab=txt2img folders=['models\\Lora'] list=0.01
thumb=0.00 desc=0.00 info=0.00 workers=12
10:00:18-639288 DEBUG Networks: type="style" items=288 subfolders=3 tab=txt2img folders=['models\\styles', 'html']
list=0.03 thumb=0.00 desc=0.00 info=0.00 workers=12
10:00:18-649503 DEBUG Networks: type="wildcards" items=0 subfolders=1 tab=txt2img folders=['models\\wildcards']
list=0.00 thumb=0.00 desc=0.00 info=0.00 workers=12
10:00:18-654759 DEBUG Networks: type="embedding" items=0 subfolders=1 tab=txt2img folders=['models\\embeddings']
list=0.00 thumb=0.00 desc=0.00 info=0.00 workers=12
10:00:18-665009 DEBUG Networks: type="vae" items=0 subfolders=1 tab=txt2img folders=['models\\VAE'] list=0.00
thumb=0.00 desc=0.00 info=0.00 workers=12
10:00:18-670417 DEBUG Networks: type="history" items=0 subfolders=1 tab=txt2img folders=[] list=0.00 thumb=0.00
desc=0.00 info=0.00 workers=12
10:00:18-914550 DEBUG UI initialize: tab=img2img
10:00:19-185281 DEBUG UI initialize: tab=control models="models\control"
10:00:19-696623 DEBUG UI initialize: tab=video
10:00:19-831099 DEBUG UI initialize: tab=process
10:00:19-887684 DEBUG UI initialize: tab=caption
10:00:20-006740 DEBUG UI initialize: tab=models
10:00:20-086335 DEBUG UI initialize: tab=gallery
10:00:20-132405 DEBUG Read: file="ui-config.json" json=0 bytes=2 time=0.000 fn=__init__:read_from_file
10:00:20-135838 DEBUG UI initialize: tab=settings
10:00:20-687363 DEBUG Settings: sections=23 settings=377/601 quicksettings=1
10:00:20-757177 DEBUG UI initialize: tab=info
10:00:20-781113 DEBUG UI initialize: tab=extensions
10:00:20-786133 INFO Extension list is empty: refresh required
10:00:21-443944 DEBUG Extension list: processed=3 installed=3 enabled=3 disabled=0 visible=3 hidden=0
10:00:21-840908 DEBUG Root paths: ['E:\\Anewforge\\SDNEXT\\sdnext', 'models']
10:00:21-964069 INFO Local URL: http://127.0.0.1:7860/
10:00:21-966035 INFO API docs: http://127.0.0.1:7860/docs
10:00:21-968030 INFO API redocs: http://127.0.0.1:7860/redocs
10:00:21-969997 DEBUG API middleware: [<class 'starlette.middleware.base.BaseHTTPMiddleware'>, <class
'starlette.middleware.gzip.GZipMiddleware'>]
10:00:21-972992 DEBUG API initialize
10:00:22-245660 DEBUG Scripts setup: time=0.462 ['XYZ Grid:0.063', 'IP Adapters:0.051', 'Prompt enhance:0.031',
'Mixture-of-Diffusers: Tile Control:0.031']
10:00:22-249738 DEBUG Model metadata: file="metadata.json" no changes
10:00:22-250736 INFO Model: autoload=True selected="waiIllustriousSDXL_v160 [a5f58eb1c3]"
10:00:22-255940 DEBUG Model requested: fn=threading.py:run:<lambda>
10:00:22-257937 DEBUG Search model: name="waiIllustriousSDXL_v160 [a5f58eb1c3]"
matched="E:\Anewforge\SDNEXT\sdnext\models\Stable-diffusion\waiIllustriousSDXL_v160.safetensor
s" type=alias
10:00:22-264529 INFO Load model: select="waiIllustriousSDXL_v160 [a5f58eb1c3]"
10:00:22-269515 INFO Autodetect model: detect="Stable Diffusion XL" class=StableDiffusionXLPipeline
file="E:\Anewforge\SDNEXT\sdnext\models\Stable-diffusion\waiIllustriousSDXL_v160.safetensors"
10:00:22-272507 DEBUG Cache clear
Progress 2.31it/s █████████ 100% 7/7 00:03 00:00 Loading pipeline components...
10:00:25-711621 DEBUG Setting model: pipeline=StableDiffusionXLPipeline config={'low_cpu_mem_usage': True,
'torch_dtype': torch.bfloat16, 'load_connected_pipeline': True, 'extract_ema': False,
'config': 'configs/sdxl', 'use_safetensors': True, 'cache_dir': 'models\\huggingface'}
10:00:25-730789 INFO Network load: type=embeddings loaded=0 skipped=0 time=0.00
10:00:25-732754 DEBUG Setting model: component=vae {'slicing': True, 'tiling': False}
10:00:25-735979 DEBUG Setting model: attention="Scaled-Dot-Product"
10:00:25-741850 INFO Offload: type=balanced op=init watermark=0.2-0.6 gpu=3.20-9.60:16.00 cpu=64.000 limit=0.00
always=[] never=[] pre=True streams=False
10:00:25-777782 DEBUG Module: name=unet cls=UNet2DConditionModel size=4.782 params=2567463684 quant=None
10:00:25-788801 DEBUG Module: name=text_encoder_2 cls=CLIPTextModelWithProjection size=1.294 params=694659840
quant=None
10:00:25-797346 DEBUG Module: name=text_encoder cls=CLIPTextModel size=0.229 params=123060480 quant=None
10:00:25-810210 DEBUG Module: name=vae cls=AutoencoderKL size=0.156 params=83653863 quant=None
10:00:25-814201 INFO Model class=StableDiffusionXLPipeline modules=4 size=6.461
10:00:25-825202 INFO Load model: family=sdxl time={'total': 3.55, 'load': 3.46} native=1024 memory={'ram':
{'total': 63.91, 'rss': 8.5, 'used': 20.3, 'free': 43.62, 'avail': 43.62, 'buffers': 0,
'cached': 0}, 'gpu': {'used': 0.18, 'total': 15.98, 'active': 0.0, 'peak': 0.0, 'retries': 0,
'oom': 0, 'swap': 0}, 'job': 'Load model'}
10:00:25-844164 DEBUG Script init: ['system-info.py:app_started=0.10']
10:00:25-846117 DEBUG Save: file="E:\Anewforge\SDNEXT\sdnext\config.json" json=20 bytes=1157 time=0.002
10:00:25-851103 INFO Startup time: total=122.18 launch=33.80 loader=33.59 installer=33.59 gradio=4.33 torch=4.22
checkpoint=3.60 libraries=2.53 diffusers=1.87 ui-extensions=1.29 ui-networks=0.43
upscalers=0.41 ui-control=0.35 extensions=0.33 ui-defaults=0.31 ui-img2img=0.22
ui-txt2img=0.21 api=0.17 ui-models=0.16
10:00:59-187617 INFO API user=None code=200 http/1.1 GET /sdapi/v1/sd-models 127.0.0.1 0.004
10:00:59-238479 INFO API user=None code=200 http/1.1 GET /sdapi/v1/version 127.0.0.1 0.001
10:00:59-276377 INFO Browser session: user=None client=127.0.0.1 agent=Mozilla/5.0 (Windows NT 10.0; Win64; x64;
rv:146.0) Gecko/20100101 Firefox/146.0
10:00:59-278372 INFO UI: ready time=10037
10:00:59-285355 INFO API user=None code=200 http/1.1 GET /sdapi/v1/start 127.0.0.1 0.014
10:00:59-363146 INFO API user=None code=200 http/1.1 GET /sdapi/v1/motd 127.0.0.1 0.001
10:00:59-387082 DEBUG UI: connected
10:00:59-388079 INFO API user=None code=200 http/1.1 GET /sdapi/v1/version 127.0.0.1 0.002
10:01:10-691988 DEBUG Sampler: "Default" cls=EulerAncestralDiscreteScheduler config={'num_train_timesteps': 1000,
'beta_start': 0.00085, 'beta_end': 0.012, 'beta_schedule': 'scaled_linear', 'trained_betas':
None, 'prediction_type': 'epsilon', 'timestep_spacing': 'trailing', 'steps_offset': 1,
'rescale_betas_zero_snr': False, 'interpolation_type': 'linear', 'use_karras_sigmas': False,
'clip_sample': False, 'sample_max_value': 1.0, 'set_alpha_to_one': False, 'skip_prk_steps':
True}
10:01:10-717921 INFO Processing modifiers: apply
10:01:10-757836 DEBUG Tokenizer: words=49408 file="configs/sdxl\tokenizer\vocab.json"
rocblaslt error: Cannot read "E:\\Anewforge\\SDNEXT\\sdnext\\venv\\Lib\\site-packages\\_rocm_sdk_libraries_gfx110X_dgpu\\bin\\hipblaslt\\library\\TensileLibrary_lazy_gfx1102.dat": No error
rocblaslt error: Could not load "E:\\Anewforge\\SDNEXT\\sdnext\\venv\\Lib\\site-packages\\_rocm_sdk_libraries_gfx110X_dgpu\\bin\\hipblaslt\\library\\TensileLibrary_lazy_gfx1102.dat"
10:01:11-605546 ERROR Prompt parser encode: CUDA error: HIPBLAS_STATUS_INVALID_VALUE when calling
`hipblasLtMatmulAlgoGetHeuristic( ltHandle, computeDesc.descriptor(), Adesc.descriptor(),
Bdesc.descriptor(), Cdesc.descriptor(), Cdesc.descriptor(), preference.descriptor(), 1,
&heuristicResult, &returnedResult)`
10:01:11-621503 INFO Base: pipeline=StableDiffusionXLPipeline task=TEXT_2_IMAGE batch=1/1x1 set={'prompt': 1,
'negative_prompt': 1, 'guidance_scale': 6, 'generator': 'cuda:[1406742169]',
'num_inference_steps': 20, 'eta': 1.0, 'guidance_rescale': 0, 'denoising_end': None,
'output_type': 'latent', 'width': 1024, 'height': 1024, 'parser': 'fixed'}
10:01:11-646437 ERROR Processing: step=base args={'prompt': ['woman, black dress'], 'negative_prompt': [''],
'guidance_scale': 6, 'generator': [<torch._C.Generator object at 0x00000218F9755D50>],
'callback_on_step_end': <function diffusers_callback at 0x00000218F1F2C040>,
'callback_on_step_end_tensor_inputs': ['latents', 'prompt_embeds', 'add_text_embeds',
'add_time_ids', 'noise_pred'], 'num_inference_steps': 20, 'eta': 1.0, 'guidance_rescale': 0,
'denoising_end': None, 'output_type': 'latent', 'width': 1024, 'height': 1024} CUDA error:
HIPBLAS_STATUS_INVALID_VALUE when calling `hipblasLtMatmulAlgoGetHeuristic( ltHandle,
computeDesc.descriptor(), Adesc.descriptor(), Bdesc.descriptor(), Cdesc.descriptor(),
Cdesc.descriptor(), preference.descriptor(), 1, &heuristicResult, &returnedResult)`
10:01:11-665387 ERROR Processing: RuntimeError
┌───────────────────────────────────────── Traceback (most recent call last) ─────────────────────────────────────────┐
│E:\Anewforge\SDNEXT\sdnext\modules\processing_diffusers.py:180 in process_base │
│ │
│ 179 │ │ │ taskid = shared.state.begin('Inference') │
│> 180 │ │ │ output = shared.sd_model(**base_args) │
│ 181 │ │ │ shared.state.end(taskid) │
│ │
│E:\Anewforge\SDNEXT\sdnext\venv\Lib\site-packages\torch\utils\_contextlib.py:120 in decorate_context │
│ │
│ 119 │ │ with ctx_factory(): │
│> 120 │ │ │ return func(*args, **kwargs) │
│ 121 │
│ │
│E:\Anewforge\SDNEXT\sdnext\venv\Lib\site-packages\diffusers\pipelines\stable_diffusion_xl\pipeline_stable_diffusion_ │
│ │
│ 1079 │ │ │ negative_pooled_prompt_embeds, │
│> 1080 │ │ ) = self.encode_prompt( │
│ 1081 │ │ │ prompt=prompt, │
│ │
│E:\Anewforge\SDNEXT\sdnext\venv\Lib\site-packages\diffusers\pipelines\stable_diffusion_xl\pipeline_stable_diffusion_ │
│ │
│ 404 │ │ │ │ │
│> 405 │ │ │ │ prompt_embeds = text_encoder(text_input_ids.to(device), output_hidden_states=True) │
│ 406 │
│ │
│E:\Anewforge\SDNEXT\sdnext\venv\Lib\site-packages\torch\nn\modules\module.py:1775 in _wrapped_call_impl │
│ │
│ 1774 │ │ else: │
│> 1775 │ │ │ return self._call_impl(*args, **kwargs) │
│ 1776 │
│ │
│E:\Anewforge\SDNEXT\sdnext\venv\Lib\site-packages\torch\nn\modules\module.py:1786 in _call_impl │
│ │
│ 1785 │ │ │ │ or _global_forward_hooks or _global_forward_pre_hooks): │
│> 1786 │ │ │ return forward_call(*args, **kwargs) │
│ 1787 │
│ │
│E:\Anewforge\SDNEXT\sdnext\venv\Lib\site-packages\accelerate\hooks.py:175 in new_forward │
│ │
│ 174 │ │ else: │
│> 175 │ │ │ output = module._old_forward(*args, **kwargs) │
│ 176 │ │ return module._hf_hook.post_forward(module, output) │
│ │
│E:\Anewforge\SDNEXT\sdnext\venv\Lib\site-packages\transformers\utils\generic.py:918 in wrapper │
│ │
│ 917 │ │ │ return_dict = return_dict_passed │
│> 918 │ │ output = func(self, *args, **kwargs) │
│ 919 │ │ if not return_dict and not isinstance(output, tuple): │
│ │
│ ... 9 frames hidden ... │
│ │
│E:\Anewforge\SDNEXT\sdnext\venv\Lib\site-packages\torch\nn\modules\module.py:1786 in _call_impl │
│ │
│ 1785 │ │ │ │ or _global_forward_hooks or _global_forward_pre_hooks): │
│> 1786 │ │ │ return forward_call(*args, **kwargs) │
│ 1787 │
│ │
│E:\Anewforge\SDNEXT\sdnext\venv\Lib\site-packages\transformers\models\clip\modeling_clip.py:397 in forward │
│ │
│ 396 │ │ hidden_states = self.layer_norm1(hidden_states) │
│> 397 │ │ hidden_states, attn_weights = self.self_attn( │
│ 398 │ │ │ hidden_states=hidden_states, │
│ │
│E:\Anewforge\SDNEXT\sdnext\venv\Lib\site-packages\torch\nn\modules\module.py:1775 in _wrapped_call_impl │
│ │
│ 1774 │ │ else: │
│> 1775 │ │ │ return self._call_impl(*args, **kwargs) │
│ 1776 │
│ │
│E:\Anewforge\SDNEXT\sdnext\venv\Lib\site-packages\torch\nn\modules\module.py:1786 in _call_impl │
│ │
│ 1785 │ │ │ │ or _global_forward_hooks or _global_forward_pre_hooks): │
│> 1786 │ │ │ return forward_call(*args, **kwargs) │
│ 1787 │
│ │
│E:\Anewforge\SDNEXT\sdnext\venv\Lib\site-packages\transformers\models\clip\modeling_clip.py:312 in forward │
│ │
│ 311 │ │ │
│> 312 │ │ queries = self.q_proj(hidden_states) │
│ 313 │ │ keys = self.k_proj(hidden_states) │
│ │
│E:\Anewforge\SDNEXT\sdnext\venv\Lib\site-packages\torch\nn\modules\module.py:1775 in _wrapped_call_impl │
│ │
│ 1774 │ │ else: │
│> 1775 │ │ │ return self._call_impl(*args, **kwargs) │
│ 1776 │
│ │
│E:\Anewforge\SDNEXT\sdnext\venv\Lib\site-packages\torch\nn\modules\module.py:1786 in _call_impl │
│ │
│ 1785 │ │ │ │ or _global_forward_hooks or _global_forward_pre_hooks): │
│> 1786 │ │ │ return forward_call(*args, **kwargs) │
│ 1787 │
│ │
│E:\Anewforge\SDNEXT\sdnext\venv\Lib\site-packages\torch\nn\modules\linear.py:134 in forward │
│ │
│ 133 │ │ """ │
│> 134 │ │ return F.linear(input, self.weight, self.bias) │
│ 135 │
└─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┘
RuntimeError: CUDA error: HIPBLAS_STATUS_INVALID_VALUE when calling `hipblasLtMatmulAlgoGetHeuristic( ltHandle, computeDesc.descriptor(), Adesc.descriptor(), Bdesc.descriptor(), Cdesc.descriptor(), Cdesc.descriptor(), preference.descriptor(), 1, &heuristicResult, &returnedResult)`
10:01:12-913169 DEBUG Search model: name="waiIllustriousSDXL_v160 [a5f58eb1c3]"
matched="E:\Anewforge\SDNEXT\sdnext\models\Stable-diffusion\waiIllustriousSDXL_v160.safetensor
s" type=alias
10:01:12-941125 DEBUG Analyzed: model="waiIllustriousSDXL_v160" type=sdxl class=StableDiffusionXLPipeline
size=6938040682 mtime="2026-01-01 17:43:50" modules=[name="vae" cls=AutoencoderKL config=True
device=cpu dtype=torch.bfloat16 params=83653863 modules=243, name="text_encoder"
cls=CLIPTextModel config=True device=cuda:0 dtype=torch.bfloat16 params=123060480 modules=152,
name="text_encoder_2" cls=CLIPTextModelWithProjection config=True device=cuda:0
dtype=torch.bfloat16 params=694659840 modules=393, name="tokenizer" cls=CLIPTokenizer
config=False, name="tokenizer_2" cls=CLIPTokenizer config=False, name="unet"
cls=UNet2DConditionModel config=True device=cpu dtype=torch.bfloat16 params=2567463684
modules=1930, name="scheduler" cls=EulerAncestralDiscreteScheduler config=True,
name="image_encoder" cls=NoneType config=False, name="feature_extractor" cls=NoneType
config=False, name="force_zeros_for_empty_prompt" cls=bool config=False]
10:01:12-964526 INFO Processing modifiers: unapply
10:01:12-986672 DEBUG Process: batch=1/1 interrupted
10:01:12-988980 INFO Processed: images=0 its=0.00 ops=['txt2img']
10:01:12-990977 DEBUG Processed: timers={'total': 3.78, 'post': 2.25, 'prompt': 0.88, 'move': 0.57}
10:01:12-995963 DEBUG Processed: memory={'ram': {'total': 63.91, 'rss': 8.54, 'used': 21.03, 'free': 42.89, 'avail':
42.89, 'buffers': 0, 'cached': 0}, 'gpu': {'used': 1.94, 'total': 15.98, 'active': 1.6,
'peak': 1.6, 'retries': 0, 'oom': 0, 'swap': 0}, 'job': ''}Backend
Diffusers
Compute
AMD ROCm
Interface
Standard
Branch
Master
Model
StableDiffusion 1.5/2.1/XL
Acknowledgements
- I have read the above and searched for existing issues
- I confirm that this is classified correctly and its not an extension issue